DBpedia Viewer - An Integrative Interface for DBpedia Leveraging the DBpedia Service Eco System Denis Lukovnikov Claus Stadler Dimitris Kontokostas University of Leipzig University of Leipzig University of Leipzig lukovnikov@informatik.uni- cstadler@informatik.uni- kontokostas@informatik.uni- leipzig.de leipzig.de leipzig.de Sebastian Hellmann Jens Lehmann University of Leipzig University of Leipzig hellmann@informatik.uni- lehmann@informatik.uni- leipzig.de leipzig.de ABSTRACT distinguishing feature of DBpedia Viewer is the Triple Ac- With the growing interest in publishing data according to tion Framework, a framework which allows to dynamically the Linked Data principles, it becomes more important to associate action types with triples. provide intuitive tools for users to view and interact with resources. The characteristics of Linked Data pose several The remainder of the article is structured as follows: Sec- challenges for user-friendly presentation of the information. tion 2 provides an overview of DBpedia and the DBpedia In this work, we present the DBpedia Viewer as one method service eco system. An overview of DBpedia Viewer is given to address this problem. The DBpedia Viewer is the new in Section 3 and Section 4 gives a more detailed view of DBpedia Linked Data user interface, which makes DBpedia the DBpedia Viewer components. Section 5 elaborates on data more accessible to non-experts while integrating the related work and we conclude with Section 6. DBpedia service eco system as well as external Linked Data services. 2. DBPEDIA DBpedia is a pioneer project in Linked Data publishing. It 1. INTRODUCTION was one of the first Linked Open Data datasets available in The Linked Data principles provide guidelines for publish- 2007 and is a hub in the Linked Open Data cloud1 . The data ing structured data in order to make it easily accessible to in DBpedia originates from Wikipedia and is extracted us- both machines and humans. In contrast to the common ing the DBpedia extraction framework. The latest DBpedia natural language representation of information on the Web, release provides data for 4.0 million entities out of which 3.2 dereferencing Linked Data resources in many cases does not millions are classified according to the DBpedia ontology.2 provide an intuitive view for humans. The fine granularity of triple-based data representation standardized by the RDF DBpedia provides different kinds of information about enti- format requires mental effort to integrate information about ties. Entities typically have types, labels, links, Linked Data the described entity, even for experts. links and textual descriptions associated with them. DBpe- dia contains links to equivalent entities in other datasets This work describes DBpedia Viewer, the new DBpedia [7] (such as YAGO) and links to the same entities in DBpedia interface, which aims to present information from DBpe- datasets for other languages (such as nl.dbpedia.org). In ad- dia in an engaging way while adhering to the Linked Data dition to the general information about entities, e.g. types principles. DBpedia Viewer integrates existing DBpedia ser- and categories, DBpedia contains properties and classes spe- vices as well as external Linked Data visualization tools to cific for particular domains. For example, the entity (of type improve user-friendliness. The DBpedia interface was orig- Person) dbpedia:Barack_Obama has a property dbo:spouse inally envisioned to serve data from DBpedia datasets, but which refers to the entity dbpedia:Michelle_Obama. could lead to a customizable framework that can easily be configured for other datasets. With the new DBpedia inter- Since the start of the DBpedia project, several tools and face, we aim to show that even a generic representation of services were developed around DBpedia. DBpedia Spot- RDF data can be user-friendly and offer relevant services. A light [9] performs Entity Linking (or Named Entity Reso- lution and Disambiguation, NERD) in text by linking en- tity mentions to DBpedia entities. DBpedia Lookup3 is an 1 Linking Open Data cloud diagram, by Richard Cyganiak and Anja Jentzsch. http://lod-cloud.net/ 2 http://blog.dbpedia.org/2013/09/17/ dbpedia-39-released-including-wider-infobox-coverage\ -additional-type-statements-and-new-yago-and-wikidata\ Copyright is held by the author/owner(s). -links/ 3 LDOW2014, April 8, 2014, Seoul, Korea. http://wiki.dbpedia.org/lookup/ additional service which allows to search for DBpedia en- 4.1 Technology Stack tities using either strings or prefixes (for auto-completion). We use the Open Source edition of the Virtuoso server6 The DBpedia mappings wiki 4 is an effort to crowdsource the as our triple store. Virtuoso has a modular architecture mappings between the Wikipedia infoboxes and the DBpe- and supports plugins. These plugins use the VSP (Virtu- dia ontology. Apart from DBpedia specific tools, external oso Server Pages) language to communicate with the triple vizualization tools have been developed that use DBpedia, store, serve web requests and provide Linked Data. The pre- such as RelFinder [4] and LodLive [2]. RelFinder explores vious DBpedia interface was built as a Virtuoso plugin and the knowledge graph to find paths between two entities. the DBpedia Viewer is built on top of the existing code. LodLive provides a visually appealing way to explore in- formation associated with an entity. Most of the new interface is client-side logic written in Java- Script. As explained above, the client-side code constructs Such tools and services exist independently, but together the page by issuing SPARQL queries to the configured end- form an eco system, which provides added value for the DB- points. Therefore, it is easy to use the new interface for pedia datasets. Our new interface integrates several tools non-Virtuoso deployments by only serving the client-side in a generic manner – partly to increase user-friendliness code. However, in this case, the old server-generated HTM- and partly to showcase the achievements in the Linked Data L/RDFa interface is not preserved. space obtained so far. For the client-side logic, we used AngularJS 7 , an open- source JavaScript framework that implements the Model 3. DBPEDIA VIEWER USER INTERFACE View Controller (MVC) paradigm for web applications. For The new DBpedia Linked Data interface (DBpedia Viewer) styling and layout, we used the CSS from Twitter Boot- co-exists with the previous interface which serves Linked strap8 . Data content to machines and browsers without JavaScript support. While we can default to the new interface when it is supported by the visiting agent, users can easily switch 4.2 Features Following we provide an elaborate discussion of the new DB- between the two interfaces. pedia Viewer features. DBpedia Viewer brings several improvements on the presen- tation of information. In addition to many cosmetic changes, 4.2.1 Pretty Box DBpedia Viewer provides improved layout choices and new The pretty box (part one of Figure 1) displays important functionalities that aim to make the UI more useful. properties of the viewed entity. There is a predefined set of facts we provide, namely: (1) a picture, (2) the title, (3) the The layout has been adapted to provide additional func- types, (4) a short description and (5) links to other resources. tionality as naturally as possible. As illustrated in Figure 1, These data are generated from the set of triples describing several new features are added, each discussed in Section 4.2. the viewed resources using predefined mappings. The DB- DBpedia Viewer is provided as open source on Github5 . pedia datasets provide most of this general information for all entities. In some cases, however, the picture or links to other resources are not available. The DBpedia Viewer 4. SYSTEM OVERVIEW does not perform automatic selection of relevant properties The architecture can be divided in three levels. The base to display. This is out of the scope of this project. Our level is the triple store, which is accessible using the SPARQL goal was to develop a UI that is customizable to configure query language. The web server and server-side code are relevant properties used to adapt the view. built on top of the database. The third level is the client- side code. In the top right corner of the pretty box (Figure 1), three icons trigger the entity actions. The currently available ac- The old interface was implemented with server-side code tions are links to alternate data representations (XML/RDF, that generated a very simple HTML/RDFa page. DBpedia n3, JSON-LD, ...) as well as links to consult the resource Viewer provides two modes: A simple HTML/RDFa view using alternative Linked Data browsers. for machines and clients without JavaScript support and a rich web interface for humans based on JavaScript. A switch 4.2.2 Search Bar between the two interfaces is also available. The rich web DBpedia Viewer provides search functionality with auto- interface is based solely on client-side code to construct the complete capabilities by re-using the DBpedia Lookup ser- web page. All SPARQL queries are executed directly from vice. DBpedia Lookup enables searching for DBpedia enti- JavaScript which minimizes the page generation time and ties using strings or provides prefix-based suggestions. reduces the load on the web server. Following, we provide an overview of DBpedia Viewer tech- 4.2.3 Language Filtering The language filtering system allows the user to choose a pre- nology stack (Section 4.1). We elaborate on the interface ferred display language. This filters all literal values based features (Section 4.2) with an emphasis on the Triple Action on the user preferences and displays only the relevant val- Framework (Section 4.3). Finally, we discuss the extensibil- ues. This feature is helpful on dbpedia.org where labels and ity of the tool(Section 4.4). 6 http://virtuoso.openlinksw.com/ 4 7 http://mappings.dbpedia.org http://angularjs.org/ 5 8 https://github.com/dbpedia/dbpedia-vad-i18n http://getbootstrap.com/ Figure 1: Screenshot of the new interface. The transparent red areas highlight new features, with the associated number in the red circle corresponding to its subsection number within Section 4.2. A quick overview: (1) pretty box, (2) search bar, (3) language switcher, (4) triple filter, (5) shortcuts, (6) preview box, (7) map and (8) triple actions. abstracts exist in 12 different languages. In the case a literal For entities having location information (latitude and longi- does not exist in the preferred language, a fallback language tude), a map is shown with its coordinates. OpenStreetMap (usually English) is chosen by default. is used for the map display. 4.2.4 Triple Filtering 4.2.8 Triple Actions Part four of Figure 1 highlights the triple filtering feature. As displayed in part 8 of (Figure 1), next to each triple, dif- Triples can be filtered using both properties and values. This ferent icons exist, each representing a different triple action. is useful for the users who quickly want to find specific prop- Triple actions are enabled using conditions on the triple. erties and values. The filtering is based on string matching Thus, the set of available actions for different triples may and supports all literal values as well as URIs. be different. When the conditions are met, the action icon is displayed next to the triple. When the user clicks on the triple action icon, the action is executed. Below is an 4.2.5 Shortcut box overview of the currently implemented user actions: The shortcut box (part 5 in Figure 1) provides anchor links to some important properties of entities. However, the list • Annotation – uses DBpedia Spotlight to annotate text. of properties is currently hardcoded and contains links to Only applicable to texts of certain length. categories, types, external links, etc... • RelFinder – links to RelFinder, where the connections (including indirect ones) between the viewed entity 4.2.6 Live Previews and the value entity can be explored. Only applica- When the user hovers over a DBpedia link (URI, ontology ble to DBpedia resources. property or class) a concise, language-filtered preview is dis- played. For entities, this preview contains a picture (if avail- • LodLive – opens the value entity with the LodLive able), the title and a short description. Part 6 of Figure 1 browser. Only applicable to DBpedia resources. shows a preview of the French Gothic architecture entity. • OpenLink Faceted Browser – view the value entity us- ing OpenLink Faceted Browser. Only applicable to 4.2.7 Maps DBpedia resources. • Wikipedia – opens the Wikipedia page associated with First, we discuss different tools that are integrated in the the value entity. Only applicable to DBpedia resources. new interface. This is followed by an overview of some Linked Data browsers. • DBpedia template mapping – links to the DBpedia mapping associated with the DBpedia template. Only applicable to DBpedia resources under the Wikipedia Integrated tools template namespace. RelFinder [4] allows users to explore connections between multiple entities in a intuitive and interactive way. Given two entities, RelFinder shows paths in the underlying RDF 4.3 Triple Action Framework graph connecting the two entities. The relationship discov- The Triple Action Framework (TAF) aims to improve the ery algorithm used in [4] is based on the original DBpedia integration of tools from the DBpedia eco system on the DB- Relationship Finder algorithm [8]. The search algorithm is pedia website. The interface maintainer can easily add new essentially a breadth-first search algorithm with several op- actions or adapt existing ones for a particular deployment. timizations for the problem. TAF allows to define a triple action with the following core semantics, (1) bind and (2) execute. DBpedia Spotlight [9] is an Entity Linking (EL) system. Given a text, the purpose of EL is to find which parts of Upon page load, for each triple, the bind method of each text refer to which entities. DBpedia Spotlight performs action is called to determine whether this action is applica- EL with DBpedia entities. The linking approach of DBpe- ble for this triple. The bind method may use any informa- dia Spotlight consists of three steps: (1) the spotting stage tion available from the triple to decide whether the action where the phrases in the text are recognized that might refer is applicable or not. For example, the DBpedia Spotlight to entities, (2) the candidate selection stage where possible annotation action should only be made available for annota- ”meanings” of spotted phrases are generated and (3) the dis- tion of textual resources, so the bind method of this action ambiguation stage where the best candidate entity is chosen checks whether the object of the triple is a string literal and as the meaning of the phrase. whether it exceeds a configurable minimum length. Another tool integrated as a triple action is LodLive [2]. The execute method of an action is called when the user LodLive is an exploratory tool that allows users to browse clicks on an action icon next to the triple. For the DBpedia Linked Data in an interactive way, using a dynamic visual Spotlight annotation action, this method uses the text in graph. Moreover, it integrates information available across the object of the triple, sends it to the DBpedia Spotlight different SPARQL endpoints. This way, it aims to showcase API for annotation and waits for a response. When the API the principles behind Linked Data. responds, the Spotlight action changes the display value of the object of the triple to show the annotations. Linked Data browsers and integrators In the actual implementation, TAF provides additional hooks, Dadzie and Rowe [3] performed a survey of tools for Linked providing more functionality to define new actions with ease. Data consumption. In their review, the authors make a Moreover, the Triple Action Framework allows to define lo- distinction between visualization (e.g. RelFinder[4]) and cally/globally stateful actions and hidden (system-level) ac- presentation (e.g. Marbles). A wide range of tools is dis- tions that execute upon binding and are not available to the cussed and a comparative study of their features is per- users. Actions of the hidden type are used to detect coordi- formed. They also make a distinction between three kinds of nates for the map and to populate the shortcut box as well users: (1) tech-users, (2) domain experts and (3) lay users. as some parts of the pretty box. One of the conclusions of their survey is that the reviewed Linked Data consumption tools are mostly oriented at tech 4.4 Extensibility users. Some of the tools discussed by Dadzie and Rowe [3] The triple action framework (TAF) promotes the extensibil- are discussed in this section. ity of the new interface. The website maintainer can quickly define new actions and add them to the interface. TAF The Marbles Linked Data browser 9 is a server-side appli- lightens adding new functionality by providing a useful ab- cation that generates HTML from Semantic Web content straction where easy access to the triple is provided and using Fresnel [10] vocabularies. Marbles is used in DBpedia displaying is already taken care of. To create a new action, Mobile [1], a location-based Linked Data browser. DBpedia one simply needs to implement the hooks with the desired Mobile shows locations available from DBpedia on a map action logic. We are looking for ways to make TAF action with information about the location. creation easier. Pubby 10 is a server-side Java application that can be con- DBpedia is distributed in many language chapters [5]. DB- figured to use a SPARQL endpoint and publish the data pedia Viewer can be deployed on all DBpedia language edi- behind it as Linked Data. It also provides a simple (static) tions by changing the configuration on the Virtuoso server. HTML user interface. The Graphity project 11 provides a When deploying to other DBpedia chapters, the functional- framework for publishing RDF data or building applications ity supported in the DBpedia Viewer depends on whether around it. LDIF [11] is a framework aiming to integrate in- this functionality is available for that dataset. 9 http://mes.github.io/marbles 10 http://wifo5-03.informatik.uni-mannheim.de/pubby/ 5. RELATED WORK 11 https://github.com/Graphity/graphity-browser formation about entities from different datasets but does not [3] Aba-Sah Dadzie and Matthew Rowe. Approaches to focus on displaying data. visualising linked data: A survey. Semantic Web, 2(2):89–124, 2011. 6. CONCLUSION AND FUTURE WORK [4] Philipp Heim, Steffen Lohmann, and Timo Stegemann. The DBpedia Viewer is a step towards a customizable frame- Interactive Relationship Discovery via the Semantic work for interactive, user-friendly presentation of Linked Web. In Proceedings of the 7th Extended Semantic Data. The original goal was a DBpedia-specific user in- Web Conference (ESWC 2010), volume 6088 of LNCS, terface that integrates some tools working with DBpedia. pages 303–317, Berlin/Heidelberg, 2010. Springer. However, the Triple Action Framework also proved to be [5] Dimitris Kontokostas, Charalampos Bratsas, Sören useful for defining system actions, which allow for greater Auer, Sebastian Hellmann, Ioannis Antoniou, and and easier customization of the interface. George Metakides. Internationalization of Linked Data: The case of the Greek DBpedia edition. Web The interface does not try to conceal the technical philos- Semantics: Science, Services and Agents on the World ophy behind Linked Data. Instead, it embraces the philos- Wide Web, 15(0):51 – 61, 2012. ophy and presents the data as it is in a visually appealing [6] Dimitris Kontokostas, Amrapali Zaveri, Sören Auer, fashion, highlighting the underlying ideas and demonstrat- and Jens Lehmann. TripleCheckMate: A Tool for ing the possibilities of the integrated tools. Crowdsourcing the Quality Assessment of Linked Data. In Proceedings of the 4th Conference on The Triple Action Framework (TAF) introduced with DB- Knowledge Engineering and Semantic Web, 2013. pedia Viewer demonstrates a method for adding interactive [7] Jens Lehmann, Robert Isele, Max Jakob, Anja functionality to Linked Data, going beyond merely serving Jentzsch, Dimitris Kontokostas, Pablo N. Mendes, RDF facts. Such ideas may not only inspire improvements Sebastian Hellmann, Mohamed Morsey, Patrick van in other Linked Data interfaces but might also evolve to Kleef, Sören Auer, and Christian Bizer. DBpedia - A a standardized framework for human interaction with data Large-scale, Multilingual Knowledge Base Extracted across the Semantic Web in the future. from Wikipedia. Semantic Web Journal, 2014. [8] Jens Lehmann, Jörg Schüppel, and Sören Auer. Future versions of DBpedia Viewer will enforce a stronger Discovering Unknown Connections - the DBpedia modularization between triples, actions and display, thus Relationship Finder. In Proceedings of the 1st following the MVC design pattern. This will generalize the Conference on Social Semantic Web (CSSW 2007), interface into a framework that can be customized for dif- volume 113 of LNI, pages 99–110. GI, 2007. ferent datasets. We also plan to add triple actions for the [9] Pablo N Mendes, Max Jakob, Andrés Garcı́a-Silva, incorporation of triple validation by the end users [6, 12]. and Christian Bizer. DBpedia Spotlight: shedding Another triple action we are investigating is the option to light on the Web of Documents. In Proceedings of the automatically import DBpedia triples into WikiData. 7th International Conference on Semantic Systems, pages 1–8. ACM, 2011. A potential area of future research is the analysis of user [10] Emmanuel Pietriga, Christian Bizer, David Karger, behavior on the interface to produce novel Entity Summa- and Ryan Lee. Fresnel: A browser-independent rization and Entity Ranking data and methods. Entity Sum- presentation vocabulary for rdf. In Isabel Cruz, Stefan marization scores can be useful for Question Answering and Decker, Dean Allemang, Chris Preist, Daniel Schwabe, Semantic Relatedness. The scores can also be used to ex- Peter Mika, Mike Uschold, and LoraM. Aroyo, editors, tend the pretty box with the most important entity-specific The Semantic Web - ISWC 2006, volume 4273 of information (e.g.: birth place for persons) and to compute a Lecture Notes in Computer Science, pages 158–171. better list of properties for the shortcut box. Springer Berlin Heidelberg, 2006. [11] Andreas Schultz, Andrea Matteini, Robert Isele, Acknowledgements Pablo N Mendes, Christian Bizer, and Christian We would like to thank Google for sponsoring this work Becker. LDIF - A Framework for Large-Scale Linked through the Google Summer of Code 2013 project.12 This Data Integration. In 21st International World Wide work was also supported by grants from the European Union’s Web Conference (WWW 2012), Developers Track, 7th Framework Programme provided for the projects LOD2 Lyon, France, 2012. (GA no. 257943) and GeoKnow (GA no. 318159). [12] Amrapali Zaveri, Dimitris Kontokostas, Mohamed A. Sherif, Lorenz Bühmann, Mohamed Morsey, Sören 7. REFERENCES Auer, and Jens Lehmann. User-driven Quality [1] Christian Becker and Christian Bizer. DBpedia Evaluation of DBpedia. In To appear in Proceedings of Mobile: A Location-Enabled Linked Data Browser. 9th International Conference on Semantic Systems, LDOW, 369, 2008. I-SEMANTICS ’13, Graz, Austria, September 4-6, [2] Diego Valerio Camarda, Silvia Mazzini, and 2013. ACM, 2013. Alessandro Antonuccio. LodLive, exploring the Web of Data. In Proceedings of the 8th International Conference on Semantic Systems, pages 197–200. ACM, 2012. 12 http://www.google-melange.com/gsoc/org2/google/ gsoc2013/dbpediaspotlight